Sensory input in the mammalian olfactory bulb arises from olfactory sensory neurons located within the nasal epithelium that project their axons to distinct networks known as glomeruli, each corresponding to a unique feature of an odorant molecule. A biological limitation of the olfactory system is the inability of odorant receptors located on olfactory sensory neurons to be perfectly selective. Odorant receptors have a preferred ligand, although they may also bind additional ligands across a spectrum of affinities. The ability of odorant receptors to bind both preferred and non-preferred ligands could obscure the brain's interpretation of the actual ligand present. I propose that this problem is addressed mainly through the activation of the local network of neurons that surround glomeruli. These cells, which include glutamatergic external tufted (ET) cells and GABAergic periglomerular (PG) cells, act as a signal detection mechanism, selectively filtering out weak signals in favor of strong signals. Aim 1 will examine the possibility that PG cells and ET cells are activated by varying degrees of sensory input, allowing for low-level olfactory input to selectively activate inhibitory PG cells, thus effectively filtering weak sensory input signals. Ai 2 explores whether PG cell activity can be modulated by local glutamate released from ET cells, which may increase the effectiveness of weak signals and produce temporal patterning in glomerular activity. Aim 3 tests the ability of the ET cell-PG cell microcircuit to control the probability of mitral cell activation and also employs a novel population analysis that allows for identification of excitatory and inhibitory neurons in the glomerular layer of the olfactory bulb. This proposal examines a model for intra-glomerular interactions in which PG cells gate glomerular activation.